Boomerang sampling, applied to a pretrained music diffusion model, creates audio variations that improve beat tracking when training data is scarce and can change instruments via text prompts.
Our results show that it is suitable for data augmentation in training a beat and downbeat detector, but more so if only limited train- ing data is available
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.SD 1years
2025 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Music Boomerang: Reusing Diffusion Models for Data Augmentation and Audio Manipulation
Boomerang sampling, applied to a pretrained music diffusion model, creates audio variations that improve beat tracking when training data is scarce and can change instruments via text prompts.